Is the gap still important?

The difference between networking and not working is one letter and a gap left to close.

1The whole history of electronic-based networking has been spent in an effort to close the gaps between people. First, the telegraph closed the gaps between towns with messages carried to and from telegraph offices. Then the telephone closed the gaps between people in towns with connections in many homes and businesses. Eventually those networks spread to interconnect continents with wired networks.

Shortly thereafter, others questioned what would happen if we used radio waves instead of wires, and another revolution started with wireless networking. Many of these networks were broadcast in nature - radio and television connected billions of people over the airwaves. Cellular phone networks proliferated to cover the Earth. The distance between one person and the rest of the world continued to drop.

But on yet another arc, as people-based networks gained in popularity someone got the idea that if we connected machines with networks, we could get more work done faster. Those machines were large computers at first but soon moved into smaller PCs.

And now, as computers (or microcontrollers) show up in just about everything, machine-to-machine or device networks have gotten both smaller and larger. Networking technology can be put into just about any device. Networks as small as a single desk or as large as an electrical grid can be formed and managed - that is, as long as all the devices that need to talk agree on how to talk together, which brings us back to my opening statement.

One letter can mess you up

Getting the right letters (and numbers, and dots, and slashes, and dashes, and so on) in the configuration commands and files can be quite entertaining, as anyone who has typed in more than a couple IP addresses and host names knows. But it's a straightforward process that can be done. If the network agrees on the protocol, it usually can be configured, devices can be found, problems can be troubleshot, and things can be made to run and kept that way.

If life was just that simple, to the point where there was only one network. Within easy reach at my desk, I have an analog phone network, an EDGE wireless phone network, USB, Ethernet, Bluetooth, Wi-Fi, DOCSIS cable, some proprietary wireless in my keyboard and mouse, and more wireless in a remote controller for my ceiling fan. I suppose this desk could tie to the rest of the house with ZigBee and power-line networks.

Each of these networks exists for a reason, doing a particular job in a certain way. And they all have different configuration characteristics. Given the complexity on my little desk alone, thinking about the world of machines in larger industrial networks makes my head hurt.

And there's more than one gap

I have no doubt you can get a few hundred or maybe even a couple thousand devices on one industrial network. But the problem I suspect many are facing in industrial networking today is like my desk times 1,000 - and there are several types of networks in play. There's some fieldbus stuff here, some Wi-Fi stuff over there, maybe some ZigBee, 802.15.4, or 6LoPAN devices around, some 900 MHz RF modems, some new gadgets with WirelessHART coming in soon Ö who knows. You'd probably like them all to talk. You'd definitely like things you don't want to talk not to be able to get in, but that's a topic for another time.

Welcome to the gap - in fact, gaps - I see. Groups of like-minded stuff talk. Some groups talk to each other better than others. But it's not one network. It's a bunch of smaller networks sort of working together, with a lot of gaps in between. This is progress, isn't it?

Maybe my assumption is wrong, and every device doesn't really need to talk to every other device, so these protocol gaps don't matter that much. Maybe the parallel idea of any person being able to dial a number to get to any other person doesn't exist in the device world. Maybe instead of having billions of industrial devices on a network, the right idea is having thousands of devices connected in millions of separate or very loosely coupled networks. If this model works for you, the gap isn't all that important. If not, we've got work left to do.

In this issue

From here, I'm going to let some experts take up the idea of why there are separate industrial networking standards, and where and how they're being used.

We'll hear from Enfora on GPS-enabled modules for asset tracking, Synapse Wireless and GreenPeak Technologies on wireless sensor network technology, Dust Networks on an interesting application of wireless sensor networks, and Innominate Security Technologies on a distributed security model.

You can reach me with your thoughts on this and other topics at Get the letters right, and we'll close the gap.